> To: [email protected] > Subject: Solaris, reboots, and buildbot > From: [EMAIL PROTECTED] (Justin Mason) > > So, the buildbot and bbmass systems have been down since Dec 25, > coincidentally since the zone was shut down and rebooted: > > jm pts/2 c-24-127-183-241 Mon Jan 2 19:33 still logged in > jm sshd c-24-127-183-241 Mon Jan 2 19:33 still logged in > felicity pts/2 peter.techtarget Mon Jan 2 15:30 - 15:54 (00:23) > felicity sshd peter.techtarget Mon Jan 2 15:30 - 15:54 (00:23) > felicity pts/2 c-24-62-33-54.hs Sun Jan 1 16:46 - 16:46 (00:00) > felicity sshd c-24-62-33-54.hs Sun Jan 1 16:46 - 16:46 (00:00) > reboot system boot Sun Dec 25 20:25 > reboot system down Sun Dec 25 06:59 > > anyone know what happened there? I don't remember hearing anything > from Mads, and that's a pretty long outage. > > The main prob is that /etc/rc3.d/S91bbmass , and /etc/rc3.d/S91buildbot > didn't seem to run -- or at least if they *were* run, the daemons didn't > stay up, and didn't log anything. > > There's a possibility that there were $PATH issues, but in case that > wasn't it: Solaris wizards (assuming there are any here): is there > anything that needs doing to get a startup script run, apart from putting > in /etc/rc3.d? I keep hearing stuff about some massively overspecified > "svc" framework, I hope that's not it. ;) > > --j.
The svc stuff is for Solaris 10 or later. Even there you can put scripts in /etc/rc*.d, but there are some restrictions. Do these scripts run if you run them manually after the system is up? Is the execute bit set? You may have to put some debugging echo lines in the scripts and then watch the console during boot. I don't think that output to standard out is logged anywhere, it just appears on the console during boot. Tom schulz Applied Dynamics Intl. [EMAIL PROTECTED]
